I used to work in an automotive paint shop. Another point of interest:
There's a limit on the number of tries you have to "repair" the paint job, because the vehicle has to go through the oven again to bake the coatings. The trips through the oven affect more than just the paint. Some panels are filled with foam; there's internal sealant and glue, and external sealant. All of those things don't really love being baked too long. For the shop I worked in, we got 2 tries to repair.
The main limit on repairs though? The windshield. Windshields are glued in with specialized urethane 'caulking'. They have to stick to the paint, and the paint has to stick to the body. That is only certified reliable with less than 5 trips through the ovens (in my shop).
---
Paint is the hardest thing to get right in the whole assembly process, because there are so many process variables. Temperature and humidity (dew point) are the biggest besides the actual paint mix, but there's also unexpected problems from every direction. Our solvent paint process (pre-2013) was very vulnerable to siloxane (silicone) lubricants. Someone in IT applied some lubricant to a label printer and ruined hundreds of paint jobs. Livestrong bracelets also ruined at least a hundred paint jobs. Someone did weekend work on the roof? Congratulations on your dust quality spill (spill means a bunch of problems).
All of the parts of the paint have to work together, and that includes things that you don't think are paint. The body is just dumb metal, right? Of course not. The metal panels have oil on them in stamping, and they MUST stay oiled right up to the point where the body takes it's first bath. If too much of the oil has evaporated, then you start getting corrosion, and in many cases you won't be able to tell until the vehicle gets to the customer.
---
Paint shops have the most complicating routing of any part of automotive assembly. The electronic quality and build data has to be synchronized with the physical process, and that data package can be updated by several different systems. Then either IT or OT (operational technology) can make a routing decision, or an individual can use a key switch to manually change routing. "Why did this vehicle go here" is not a straightforward question to answer sometimes.
> The metal panels have oil on them in stamping, and they MUST stay oiled right up to the point where the body takes it's first bath.
I toured an assembly line a few years ago. The guide told us that in the 1970s they had teams of people watching iron/steel prices, when they thought the prices where at the low they made a massive order of all iron that factory would need for the next few years. When the factory when to just-in-time (sometime in the 1980s or 1990s) they couldn't do that anymore - but they also got to get rid of all the equipment to de-rust all that stored iron: since every raw material was shipped out as a final assembly in a week there was no time for significant rust to form and so they avoided a lot of problems.
I don't work in that area, but I suspect that is done. Auto makers also have the ability to order months in advance so they can yet discounts because the steel makers know what orders they need to fill each week instead of a lot of overtime to fill a large order.
Thisbis done, of course. Risk with securing long-term prices so is that those prices might be higher than short term ones for the same future period. So usually, admittedly it is a while I did raw material procurement and aerospace is different, you use a combination of both.
You can still get squeezed so either way. But that has close to nothing to do with JIT, JIT just concerns the delivery bit (over simplified of course).
This is fascinating, thanks. Can you elaborate more on how the label printer and Livestrong bracelets ruined paint jobs? Was it something they emitted through the air? How would that be in concentrations to affect the paint job, what would happen to the paint?
Silicone has very low surface tension, imagine water on a non stick pan. If paint encounters a droplet of silicone it will shrink back on itself and form a hole around the contamination.
That sounds fascinating and complicated. I scratched my car a few years ago and decided I wanted to DIY a chameleon paint job instead of taking it to the shop (the car is worth less than the paint job if I pay retail). I’ve been slowly acquiring all the gear I need like a forced air respirator and sprayers.
I was wondering how realistic that really is to DIY or if I’m kidding myself. Sound like even if I rented a stall in a paint shop for a few weeks, I won’t have the time to finesse the technique to get a good end product, especially with how much more finicky I understand the chameleon paint to be.
My understanding about chameleon paint is that it's very hard to match if it's not all painted at once. I think the difficulty may depend on your expectations for the final product.
Yep from what I’ve read I’ll have to do the full body in a single shot after spending several hundred man hours sanding down the old paint job and prepping the surface. I’m hoping in Southern California’s weather I won’t have much problem with rust between the prep and actual painting.
I don’t care so much about getting the matching perfect - “it was my first time” is a perfectly cromulent excuse in my social circles - but I do want the paint job to be good enough so that I’m not redoing it every year or two (I hate sanding). What should I exepct?
I don't think I can help much more, sorry. The shop I worked in was completely automated and worked on new builds. In most ways, it's much harder to do a one-off repaint of an assembled vehicle.
Only advise I can give is to get some cheap paint and some kind of panels and start practicing.
Sometimes the local tech school will give you a deal on painting. Students do the work and it will have flaws, but better than you can do without practice
Thank you for this comment. I enjoy hearing about intricacies of engineering processes-- particularly those that sound "easy" on the surface.
This gives me a deeper appreciation for cars with plastic body panels. I also wonder if DeLorean Motor Company took any of this into account when they designed the DMC-12 with stainless steel body panels.
Plastic panels are a whole bucket of issues in themselves. As far as I know, the color matching plastic panels are painted using a different process. That means the the front and rear fascia (bumper) of your vehicle are painted in a different plant (usually) from the rest of the vehicle. The colors have to match to a very high degree.
And yes, that's a big reason why they went with stainless. Stainless steel does corrode, just much less visibly than regular steel. It also has different physical properties like work hardening etc. It's also VERY expensive. Salty napkin math says $10,000 more for the Cybertruck.
> Plastic panels are a whole bucket of issues in themselves. As far as I know, the color matching plastic panels are painted using a different process.
I was thinking about "molded-in color" panels like the old Pontiac Fiero. That makes sense, though, re: painted plastic panels. They'd be a whole different ballgame and there's, I'm sure, interaction between the paint chemistry and the plastic chemistry that's probably pretty hairy.
Color matching seems hard, I almost think it is worth making the distinction clear and not even trying if you have different materials. I've watched my girlfriend hunt down matching dye lots of yarns to finish gifts when she runs out too early.
I bought 2 pairs of the same shorts a couple years ago, one in blue and one in olive. Both were used almost interchangeably in terms of exposure to sun, dirt, washing, and salt water. But on the green ones a section of the waistband is clearly differently colored where on the blue the colors still match.
The DMC-12 had stainless steel body panels, but it had a plain ol' carbon steel frame (designed by Lotus and similar in design to the Esprit) with an epoxy coating, and they have NOT fared well over the years. You can buy a new-build stainless frame for it, it only costs as much as a new regular car.
Same with old Land Rovers, people belive they don't rust being all aluminium. Chassis are steel, and those rot. Plus some other bits and pieces. New galvanised chassis are available, getting the body mounted is the fun bit so.
As an old-timer, I can tell you that modern automotive paint is off the charts gorgeous end of a quality that would have looked like alien technology in the
1970s. One color you will see frequently is a metallic red that consists of the red pigment with fine metal particles that give it a deep glimmer. 30 years ago the only way you could get that was a very expensive coding process that required 10, 12, 20 layers of paint. Now anyone can get it on their Corolla for little or no markup.
Another piece of trivia is that Fender, the guitar company, has often used automotive paints for its guitars. The reason is not just availability and durability, but that it accurately tracks recent color trends.
Fender only followed car color trends in the late 50s and early 60s. After that they went their own way (including natural wood bodies and many "interesting" color choices of the 70s.)
Most Fender guitars sold today are still in those "custom" Duco colors from the late 50s and early 60s. (The standard Stratocaster was a burst color and the standard Telecaster was butterscotch or blonde. Any of the Duco car colors back then was a "custom" option for an additional charge.)
I interviewed their #2 guy for a radio show a few years ago and he's the one who gave me that info. In fact, they follow lots of other color trends--right up to and including women's shoes (not joking). I called him out on it and he said, "Do you have any better ideas?" and that shut me up really fast.
Mazda Soul Red is what's called a "Candy Color“. It's achieved with a translucent red layer applied over a base color layer.
It's rare for an OEM to offer a color like this because the thickness the coating is applied changes the color and It's much more difficult to achieve a consistent color.
There's so much more detail you can go into in this stuff.
Before the Model T assembly line in 1916, most manufacturers did something like 6 or 7 coats by hand because it was such thin, slow drying varnish. And to get that nice shiny enamel surface we are used to with hand brushes, they had to alternate the direction of the brushstrokes between every coat as well as sand it. In addition to time, this is a huge QC issue.
Part of the magic of the all-black Model T is that you can just dip parts that can be dipped, Japan parts that can be Japanned, and pour paint on parts that cannot. At final assembly they will all be matched up together without any mismatching paint jobs.
And we cannot understate how magical the introduction of asphaltum and naptha were to the paints. This was still pretty new technology then and the precursor to our modern petrochemical derived paints and even plastics. You could get thicker coats that dried faster. So paint could just be poured on and the surface tension would smooth the surface perfectly. And they were more durable because they produced a more flexible coat.
One of the innovations of the Dodge Brothers (that Ford later copied) was that by using more metal body components, not only do you save on labor (working with wood is much more skilled than running a metal press) but that metal can withstand the much faster oven-baked finishes.
The paint shop at a modern car assembly plant can be 50% of the square footage under the roof, and cost 50% of the outlay, so half a billion for a new one at least.
The change from VOC-borne to water-based paints was also a big deal for the environment. The lacquers did a good job, but imagine living in a neighborhood where several tons of toluene or acetone was being evaporated into your air every hour. In places like LA, this process was a huge contributor to smog in the 1960s/70s.
I used to work in an automotive paint shop. When our exhaust scrubber (burner) went down, we had to stop painting cars. We had a 'budget' of VOC exhaust per year. I don't know when that went in, I imagine in the 80's or 90's.
We switched to water based paints in 2013 and that really helped with the VOC.
My neighbour was a car painter for decades he died of dementia. My Dad and grandfather had a house painting business in the 1960s/1970s using leaded paint for most of that time. Both Dad and my grandfather died of lung diseases.
Acetone is not a VOC.
Also, Don't confuse VOC, which is about air pollution, with toxicity, which is not. Toluene is a very slow evaporating VOC that is toxic. Not all VOC are toxic and not all toxic solvents used in coatings are VOCs.
Most 2k water based urethanes with isocyanate hardeners are ~zero VOC (and colorless + odorless at less than harmful concentrations) but much more toxic to those spraying and within inhalation range. Thankfully we are moving to iso free formulations.
It happens that the isocyanate reacts with moisture in air fast enough that it won't harm many others but again, VOC != Toxicity.
I would rather have non toxic air pollution than colorless odorless toxic gas, and there are plenty of examples of VOC regulation generating the second.
It's one of the fun lies that paint companies sell (0 voc = safer for you. Non VOC is good for environment in general but tells you nothing about safety)
Acetone is absolutely a volatile organic compound (VOC). It might not be toxic, but it's an organic solvent with a high vapor pressure at relatively low temperatures.
Scientifically, yes, but it is exempt from almost all regulation as a VOC.
Since the definition here is the regulatory one, it is not a VOC as we're talking about.
It is the main carrier in almost all solvent borne VOC compliant formulations of coatings
As a result, for example, roughly all solvent borne wood coatings that are VOC compliant use acetone these days.
This has fun effects.
They used to use, mostly (but it varies a bit), n-butyl acetate.
Of course, acetone flashes off way too quickly, making it almost impossible to get coatings to level and flow properly. So the real result is that the coatings are made higher solids and then get reduced by the user with a significant amount of VOC (toluene, NBA) so they are usable again.
So sold VOC compliant but not really in practice, and everyone looks the other way because they don't really generate enough pollution to be worth cracking down[1].
[1] Moving to water borne is expensive and tricky. For example:
They mostly spray conversion varnish which is cheap and easy, and resistant to almost all chemicals. Consumers want cabinets/etc that don't get stained when you get ketchup or wine on them.
In water borne, they'd have to spray 2k urethane to get the same resistance, but doing so safely requires supplied air setups that are quite expensive, at least until iso-free formulations get better.
On the other hand, at least the early water-base painted cars were prone to premature rusting, which has its own (distributed) ecological impact. My W210 Mercedes was mechanically great the day I sent it to the scrapper for excessive rust through on the rockers and underbody structure. Hundreds of thousands of car met such an earlier than otherwise demise.
Glad you liked yours; the 210s and 220s and the original MLs were the nadir of Mercedes value-engineering when it came to many things, and rustproofing was certainly one of them.
What? The W210 is the "Toyota Camry" of Mercedes-Benz. One of their most reliable cars. Yes, however, once rust got a start you could almost watch them disintegrate. Toyotas of the 1980s and 1990s were similar in that regard.
The M112 18-valve V6 is not really considered the pinnacle of M-B gas engine tech; and the interior is certainly not what lots of people expected from a Benz, certainly if you had just stepped out of a '94 500E W124. The body when prepped right is comfortable and sturdy tho. I'm sure it was heavily used in much of the world as a taxi, like a Camry (except with an M-B turbodiesel).
A good endorsement of it is that the Chrysler LX cars, which shared more than a few W210/211 bits, have been on sale with only cosmetic updates since 2005...
I had the OM606A, which was very well-regarded overall and mine went to 211K miles with no drama and the occasional glow plug was the only unscheduled maintenance I experienced in almost a decade of my ownership.
Interesting. It's still difficult to purchase in CA. Used to be sold at Home Depot, now it's "Acetone replacement". I have to go to specialty suppliers.
And Tesla is famous for trying things evrybody else knows are either stupid or just don't work economically when it comes to manufacturing cars. They sell it pretty well so, usually as some groundbreaking innovation of sorts, one that miraculiosly justifies the over hyped shares and cars.
Not that Teslas are bad cars, far from it, but tue delta between perception and substance is larger than anything German "Premium" OEMs could have ever dreamed of.
>They all use electrodeposition and most of them use robots to do most of the car spraying. They are gradually beginning to explore replacing their gas ovens with electric ones to reduce the carbon emissions at plants.
Toyota's electrodeposition and prep process can be seen here:
Pretty cool. For the next step the sprayers don't work like a normal spray gun, instead of pressure they use a spinning contraption to mist the paint and static electricity as a kind of terminal guidance to get over 90% of the paint onto the object being painted.
For those of us who would like a little more depth in the history and chemistry of automotive paints (and more images/videos too), there's a great video essay by New Mind on youtube:
these innovations spread back into the normal paint world, I've come across a number of things about modern paint that show it was first developed for cars (and mostly GM/Ford). Automobile primer is great for covering 3d prints.
> By the early 1920s, two-toned cars were the rage, and that created a major headache for the automotive industry. To craft this duo-tone look, one portion of the car had to be masked off while the other was painted. The problem was nobody knew how to do this well. So automakers and auto body shops improvised. They glued old newspapers to the body and windows with library pastes, homemade glues or surgical adhesive tape. This helped create a sharp demarcation between the two colors, but the adhesives stuck so firmly that trying to remove them often ruined the paint job.
> At the time, 3M primarily manufactured sandpaper and other abrasives. One of Dick Drew's jobs was taking samples of waterproof sandpaper (another 3M invention) to nearby auto body shops for testing. One morning in 1923 he walked into one of these shops and overheard "choicest profanity I'd ever known." Yet another paint job was botched and the worker who'd done it was furious. Drew had seen this occur on many other visits, but this time he spoke up. He could, he said, produce a tape that would end the painter's torment.
Might be a little OT but I was hoping this would include more discussion of the contemporary interplay between aesthetics and paint chemistry / technology.
I know there are some interesting stories behind titanium dioxide production, the use of which is critical to making paints vibrant and opaque [1].
Further I have a pet theory that the recent trend in 'flat' automotive paints (without the apparent depth and larger suspended metallic flakes common in decades prior) has been enabled by some kind of technical advance as well, not just a change in consumer tastes, although I have no evidence to support the idea [2]
>Further I have a pet theory that the recent trend in 'flat' automotive paints (without the apparent depth and larger suspended metallic flakes common in decades prior) has been enabled by some kind of technical advance as well, not just a change in consumer tastes, although I have no evidence to support the idea.
I've wondered about this as well. One theory I've had is that older paints used lead based pigments which gave a glossier finish, but are no longer used. I also do not have any concrete evidence to support this aside from pointing out that automotive paint was exempted from the 1978 (US) lead paint ban.
In the mid seventies I painted my newly built harpsichord in three colors with alkyd paint. Six coats with sanding between. It took about a month. After the final coats, it took several grades of wet sanding followed by rubbing compound.
Then there's composite gliders. Gelcoat is sprayed into molds before layers of fiberglass, Kevlar, or carbon fiber alternated with adhesive is laid into the mold.
Gelcoat durability is problematic. Eventually you have to sand WITHOUT damaging base layers before repainting.
Composite adhesives can be toxic. I know a canoe builder and a glider repair engineer who died prematurely – and two harpsichord builders.
> These days, where we use enormous ovens to cure and dry car coatings, it is comfortably the most energy intensive and carbon intensive parts of making a car.
I guess so if you don't count what it takes to make aluminum and steel.
It would be better phrased as "most ... intensive part of an automotive assembly plant". Generally, the only big thing that assembly plants actually manufacture is the body (including painting the body)
Once production was shut down at my plant because it was too cold and the natural gas company said they had to use the gas for residential customers.
a friend (a paint engineer) believes there are massive energy savings to be had from the painting (baking). He was shocked to see how wasteful the process is in automotive plants compared to his other clients. All because the paint has to be perfect and last long
I believe it's a myth. It was most likely due to efficiency and cost reasons. Black was cheaper than other pigments and was less likely to show imperfections.
Great article, I agree that "the influence of the chemical industry on modern products is astoundingly important, yet it’s one of the least appreciated and most rarely told stories in the modern industrial pantheon."
Can't believe I'm paying $400 for white, but even with that the overall price beats what I could get out of another dealer 100 miles away where they had a deep crystal blue on the lot. Beats paying thousands of dollars of markups for literally nothing.
Then what's boring to you is clearly tasteful to the 80% of people buying them.
I personally like to express my personality in other ways. The last thing I want is a car screaming some garish hue. Especially since I can do things like match clothes to my mood and the season, but I can't get my car repainted every day or every three months.
It's not even really up to customers today, most people (me included) will take whatever color is available as long as it's not hideous. That's how I'm ending up with a white car even though I'm not enthusiastic about it. I guess manufacturers know that, so it's no problem for them to keep cranking out the whites and grays.
Was talking to a Mazda salesperson recently and he was complaining that the first five or six of the new CX-90s they were being allocated are all white. Though looking at their website they have some red and dark blue ones on the way now, so maybe that was just the first production run.
No, you're projecting "boring" onto it. That's how you feel about what they're buying, not how they feel.
I stand by my original point that it's tasteful. Classic shades of white, gray, silver, and black are in good taste, which is why people buy them. The buyers aren't saying "I want a boring color", they're saying "I want a tasteful shade".
Bach's fugues are tasteful, french cuisine is good taste.
White and grey cars are just boring, the definition of good taste for the unimaginative petite bourgeoisie, the hallmark of conformity and middle-class sensitivity.
I have a grey car, but I am fully aware this is the case just because I am a terminally boring middle-class denizen, mass-produced, and absolutely non-remarkable for history.
McDonald's is "where the most consumer demand is" but that doesn't make it interesting food. White on a car is "good enough"; people have other priorities than getting the car in their favourite colour.
> Then what's boring to you is clearly tasteful to the 80% of people buying them.
Apparently this has to do with how the economy is doing. When times are tough, people like boring car colors. Something about ease of resale and broad palatability in car dealerships.
There's a limit on the number of tries you have to "repair" the paint job, because the vehicle has to go through the oven again to bake the coatings. The trips through the oven affect more than just the paint. Some panels are filled with foam; there's internal sealant and glue, and external sealant. All of those things don't really love being baked too long. For the shop I worked in, we got 2 tries to repair.
The main limit on repairs though? The windshield. Windshields are glued in with specialized urethane 'caulking'. They have to stick to the paint, and the paint has to stick to the body. That is only certified reliable with less than 5 trips through the ovens (in my shop).
---
Paint is the hardest thing to get right in the whole assembly process, because there are so many process variables. Temperature and humidity (dew point) are the biggest besides the actual paint mix, but there's also unexpected problems from every direction. Our solvent paint process (pre-2013) was very vulnerable to siloxane (silicone) lubricants. Someone in IT applied some lubricant to a label printer and ruined hundreds of paint jobs. Livestrong bracelets also ruined at least a hundred paint jobs. Someone did weekend work on the roof? Congratulations on your dust quality spill (spill means a bunch of problems).
All of the parts of the paint have to work together, and that includes things that you don't think are paint. The body is just dumb metal, right? Of course not. The metal panels have oil on them in stamping, and they MUST stay oiled right up to the point where the body takes it's first bath. If too much of the oil has evaporated, then you start getting corrosion, and in many cases you won't be able to tell until the vehicle gets to the customer.
---
Paint shops have the most complicating routing of any part of automotive assembly. The electronic quality and build data has to be synchronized with the physical process, and that data package can be updated by several different systems. Then either IT or OT (operational technology) can make a routing decision, or an individual can use a key switch to manually change routing. "Why did this vehicle go here" is not a straightforward question to answer sometimes.